Lamont-Doherty Geological Observatory of Columbia University, Palisades, NY10964, U.S.A. and URA CNRS 327, Institut Des Sciences

IS CAUSAL ECOLOGICAL BIOGEOGRAPHY A PROGRESSIVE RESEARCH PROGRAM?

Denis-Didier Rousseau Lamont-Doherty Geological Observatory of Columbia University, Palisades, NY10964, U.S.A. and URA CNRS 327, Institut des Sciences de l'Evolution, Universit~ Montpellier II, Place Eugdne Bataillon, 34095 Montpellier Cddex 5, France

Various spacial and temporal time scales need to be taken into account in the interpretation of biological events. Omission of these aspects can lead to purely descriptive studies of changes in faunal assemblages, which are severely limited in scope. Terrestrial molluscs, which are very sensitive to environmental and climatic changes, provide a good example to illustrate this problem. Using an ecological biogeographic approach, a review of mollusc assemblages from holsteinian tufas in western Europe is presented. The analysis permits the definition of the Lyrodiscus biome, which corresponds t ° a temperate mixed forest community which developed under a particularly warm oceanic climate, and never occurred in Western Europe since that time.

are to recognize the diversity of life and to establish the INTRODUCTION hierarchical organization of living organisms. The principle has been much used by botanists and some The new concept of biogeography, by Croizat (1958), palaeontologists, and it has allowed the recognition of associates the evolution of the Earth with that of life. floral and faunal provinces. But within it, two different Croizat replaces a purely descriptive biogeographical concepts can be distinguished (Table 1): ecological approach with a dynamic, analytical, causal process. (ecological descriptive biogeography) and phenetic Biogeographical works applying Croizat's methodology (phenetic descriptive biogeography). (pan-biogeography), or others inspired by it (e.g. First, ecological descriptive biogeography. The pro- vicariance biogeography -- Rosen, 1978) are numer- cess is eclectic and there are two approaches. The ous, but they favour the historical causal approach former makes an inventory of the communities, ana- which led to the recognition and to the analysis of lyzed against environmental controls, and mapped. biotas (historical biological communities). This paper This results in a hierarchical organization of communi- attempts to see if it may be possible to conceive a causal ties (a synecological approach). The latter makes an biogeography during a shorter period, at the scale of inventory of the taxa, case by case, and then maps 'ecological time', and to see if it is possible to integrate them. Their distribution is also analyzed against en- this approach with others. To adapt the title of the vironmental controls. This results in an identification of paper by Craw and Weston (1984) to the present morpho-functional stresses in relation to ecological discussion: is causal ecological (palaeo)biogeography conditions (autecological approach) (Table 1). possible and can it underpin research methodology? In southern and ~vestern Finland, Aho (1984) ex- Hitherto examples of such an approach based on plained the local diversity of freshwater gastropods by invertebrates, and more precisely land snails, are rare variation in water quality. In another context, Shikov despite the endeavour of Solem (1984). Nevertheless, (1984) demonstrated that human activities in the because they are well adapted to their environment, central part of European U.S.S.R. have influenced the land snails are appropriate for illustrating ecological distribution of land snails. The decrease in forest cover (palaeo)biogeography. Examples are presented from leads to some extinction of shade-demanding taxa and the global review of malacology by Solem and van to the spread of snails adapted to open habitats. Bruggen (1984). Second, phenetic biogeography may be quantitative or qualitative. In the former case, similarity indices are AlMS AND SCOPES OF BIOGEOGRAPHY calculated to compare data, to determine a hierarchical organization of distributions, and to determine An analysis of the biogeographical literature, shows (bio)geographical units. In the latter the census of taxa two major philosophical approaches: one is essentially distributions is mapped, and leads to a definition of descriptive, the other causal. A brief review of biogeo- (bio)geographical areas (Table 1). graphy is desirable to distinguish between them, using Andr6 (1984) described 829 stations in Mediterra- the summary of Dommergues and Marchand (in press). nean Spain and France. Computer analysis of 125 species allowed him to distinguish thirteen geographical Descriptive Biogeography groups on the basis of biotic and abiotic variables such Descriptive biogeography is the approach that was as: provinces or departments, chorological and ecologi- widely used by Darwin and his followers. Its main aims cal sectors and vegetation series.

593 594 D.-D. Rousseau

"FABLE 1. Synthetic description of biogeography from the approaches to the aims of studies. Comparison between the descriptive and thc causal biogeographies (modified from Dommergues and Marchand, 1988)

DESCRIPTIVE BIOGEOGRAPHY CAUSAL OR EVOLUTIVE BIOGEOGRAPHY

PHENETIC B. ECOLOGICAL B. ECOLOGICALB. HISTORICAL B.

kPPROACH no ecological or synecological autecological morpho-functlonal narrative analytical historical approach

=ROCES Quantitative Qualitative Eclectic Eclectic ] Hypothetico-deductive Eclectic Hypothetico- -SES deductive

METHODS automatical Mapping and Mappingand Mapping and inventory of taxa with similar morphc Inventory of taxa connected treatments inventory inventory inventory -functional adaptations (morphology, similarity of index taxa of communities of taxa case biology, physiology, ecology) by eclectic by eclectic or indexes by case systematic cladistic systema expert systems tic

ANALYTIC Comparison analyses of distribution of distribution of Hierarchical organization and compa- study case by compared study WAY and I~cal congruent communities in taxa in rison with the environmental data case of the pat of the patterns o ;of distributions distributions relation with environmental ~)jsoersal cen. distribution. stresses ter hypothesis Vicariant hypothesis

AIM OF Comparison Definition of Hierarchical or- Research of the BIOME$ RECOGNITION BIOTAS RECOGNITION ;TUDIES be~een(bio) (bio)geogra- ganization of the morphofunctio, recognition and definition of commum recognition and definition of com- geographical phical areas communities and nal stresses in ties with a similar morpho.functlonal munities with a similar geological units the ecological relation with adaptations and evolutionary history stresses ecological conditions spatio-temporal fac- partly by vica only by vicarian tor taken into accoun~ cariant and di,' events :pan- or persal events phylogenetic biogeography

These concepts correspond to a special approach controls, such as environment, climatology, or connec- which does not take into account the temporal factor tions between land masses, which may be confirmed by but favours an instantaneous ecological one. Such other earth sciences (Table 1). works have always had a descriptive hard core, but in This approach is well developed in a recent faunistic the view of Craw and Weston (1984) they are not an and systematic review by Solem (1984). He proposed a appropriate basis for research. generalized world model of land snail diversity and abundance. First he cited the example of the highest Causal or Evolutionary Biogeography diversity of land snails in the mid-North island of New Unlike the methods previously discussed, this Zealand and proposed arguments to explain such approach seeks to establish the distribution of living diversity. Then using that example, he attempted to organisms, but also to understand it from an historical explain the diversity of land snails in the major recent point of view, proposing hypotheses which may be biomes. He illustrated the temporal aspect of diversity confirmed by data provided by other sciences: for with, for example, the Miocene of Kenya and the example tectonics or climatology. Data sets can be Pleistocene of Europe. considered according to an ecological (Blondel, 1986) Second, historical causal biogeography. The or historical time perspective (Table 1). approach is either narrative or analytical, that is to say First, ecological causal biogeography. Although the that the process is eclectic in the former case and approach is morpho-functional, the process can be an hypothetico-deductive in the latter. Differentiation eclectic or hypothetical-deductive one. The methods exists in the method used: in the former case, an are similar: the establishment of an inventory of taxa inventory of taxa is related to Simpsonian systematics; showing one or more similar morpho-functional adap- in the latter, the taxa are related to phylogenetic tations (morphology, biology, physiology, ecology), systematics. The analysis by the narrative process which are compared with environmental data. The aim consists of a case by case study of the patterns of in the former case (eclectic) is to recognize and define distribution compared with a dispersal centre hypo- communities and their variations through time accord- thesis. In the analytical processes, the analysis is made ing to known data about the environment. In the latter by a congruent study of the patterns of distribution with case, the environment does not explain the communi- the vicariance hypothesis. In the two cases the result is ties. Using spatial and temporal factors, the aim is to the recognition of biotas defined in the former case determine biomes (ecological biogeographical com- partly by vicariance and dispersal, and in the latter case munities), and afterwards to propose hypotheses about essentially by vicariant events, in accordance with the Causal Ecological Biogeography 595 concept of pan-biogeography of Croizat (1958), or ing with a climatic sector or area. phylogenetic biogeography of Rosen (1978) and his The question of the association of plant and animal followers (Table 1). groupings needs addressing. A group of vertebrates Meier-Brook (1984) reviewed the tribe Planorbini of could easily be found, during a migratory cycle, among the Planorbidae, on the basis of cladistic analysis: the two or more distinct vegetational associations. On the distribution of the freshwater pulmonate gastropods is other hand, with their low mobility, small animals like explained by vicariant events at different levels of land snails constitute communities superimposed on family, species, genus or sub-genus. In another plant associations, and like them are characterized by a example, Gittenberger (1984) revised the land snails of dominant species. the sub-family Chondrinidae, and gave prominence to A further question is, how is a biome integrated in a two patterns of distribution in the western Palaearctic biota? A biota represents a biological superstructure at region. In this example members of Granaria show the geological time scale (Fig. 1). It groups together evidence of vicariant patterns, while species of Chon- phylogenetic systems having a closely similar history, drina indicate a dispersal pattern; other genera show a and in which the pattern of distribution is identical. combination of vicariant and dispersal events. Within a biota, it is thus possible to determine and to If we consider causal ecological and causal historical characterize one or several sub-systems, each appar- biogeographies, we notice that the differentiation ently less complex, which are instantaneous because between biotas and biomes reflects the divergence of their existence is brief (Fig. 1). Biomes can be exclusive integration levels between, on one hand, extensive for a long time. In such cases, the biome merges into areas over a relatively long time scale (historical the biota, as is probably the case with insular communi- communities: biotas), and on the other hand, 'local' ties. In another way, the distinction between biota and situations over an instantaneous geological time scale biome expresses the difference between the ecological (ecological communities: biomes) (Fig. 1). and the historical weights which are characteristic of From the definition of Lrmre (1967), following the scale of the approach (Fig. 1), At the regional scale, Clements (1949), a biome is a biogeographical entity ecological factors are always dominant. But if the scale characterized by biological individuals, and the domi- is changed, for example to a continental one, historical nant species which constitute it. Therefore, the biome factors become dominant. Ecological factors have less concept defines a major biological community which is importance at this scale. On the other hand, all the in equilibrium (that is to say at its climax), and is history of the continent, and consequently of the composed of plants and animals having homologous biological elements, may become important in explain- morpho-functional adaptations (Table 1). In other ing the observed distributions (Fig. 1). words a biome can be understood as the association, for The applications of these concepts in palaeontology a determined geographical entity, of the plant and are relatively new (Janvier, 1982, 1986) but they are animal populations adapted to particular ecological worth extending. Nevertheless, it is nearly always the conditions, but nevertheless susceptible to small spatial historical approaches which are developed, that is to variations. Thus, the ecological community can be say those corresponding to an evolutionary time-scale. associated with an ecological sector or area correspond- Because of the relative precision of its records, with

~TO~ (mANG~8) ...... ~r~ 106-z... Bs ~ wS s e s S Ds B /°

Be ° 4,s ~ s~ B

...... -a~-- - - 10 3-4

FOPtIl,h2'ION ~ ,s t-~f.."I • , !

s S , • - "A s ! I ~ " ~ "-~lt~ I BS I I TIgMPOIALIFAC'FOR $ -" -")'e" "a • . .I • - m ! 1~_.~31e.- B_ .,o0.~, ! I .** lyears} v i II / I i I I COIiO/UNI'I'Y ---,---a-- , _- s , ." , , ~_ _ _ aklO , .', ' ~! -l-O'lO | B* ! , ~10 tmOMIg, -" , ' ...... J,~ I

P ~1-_. !

• 8P.4C1~ 11'ACTOR (bm2l FIG. 1. Spatio-temporal representation of the main biological assemblagesaccording to the evolutionary changes. The dotted lines indicate the limit that each entity can move over or under the proposed boundary as expressed by the exposure of 10. 596 D.-D. Rousseau numerous horizons yielding high chronostratigraphical resolution, the Quaternary is a period for this analysis. Among the invertebrates, which are susceptible to this method, Pleistocene land snails appear to be appropriate because they show less evolution than other animal groups, such as, for example the rodents, during this period.

THE FOREST MALACOLOGICAL ASSOCIATIONS OF THE 'HOLSTEINIAN' TUFAS: AN EXAMPLE OF THE BIOME CONCEPT

The Problem The climatic optima of interglacials in central Europe are characterized by a malacofauna which is rich in thermophilous and forest species, whose present distribution is south-eastern Europe: Helicigona banatica (south-Carpathians), Soosia diodonta (north-Balkans), lphigena densetriata (east-Alpine), Aegopinella ress- manni (southeastern-Alpine), and Mastus bielzi (east- Carpathians). In addition, some taxa belonging to these FIG. 2. Map of west and central Europe with shading showing land malacofaunas are fossil, such as Helicigona capeki, or over 200 m. The arrows show the main emigration routes in they have radically changed their distribution, like interglacials (arrows: oceanic coast, Sarne-Rhone and Danubian Gastrocopta theeli (presently east-Asiatic) (Lozek, corridors). The location of the sites in the text is indicated by the numbers: 1-Icklingham,2-Hitchin, 3-Arrest, 4-St Pierre-16s-Elbeuf, 1964). This H. banatica malacological association 5-Vernon. 6-La Celle-sous-Moret. (Lozek, op. cit.) did not come farther west than southern and central Germany in the Middle Pleisto- cene (Lozek, 1969; Kukla, 1977). In western Europe, taxa occur in some interglacials Materials which have modified their distribution as compared In Europe, two pleniglacial biomes have been recog- with the present for example: Ruthenica filograna ('last nized (Rousseau, 1986). They represent two ecological interglacial' in Burgundy; Puissrgur, 1976 -- now classes: cold steppe and tundra. The latter is charac- mainly in central Europe), Aegopis verticillus ('third' terized by a richer malacological association with interglacial before the Holocene in Normandy and in greater specific diversity. Analysis of the Achenheim Alsace; Puissrgur, 1978; Lautridou et al., 1985, 1986). deposits (Rousseau, 1986, 1987) has demonstrated that These species do not indicate the southern component during the last five interglacial-glacial cycles, the of the central-European species previously cited. In glacial phases in west and central Europe have been England, numerous taxa, are found in interglacial characterized by the sharing of only one biome or by deposits, but they always indicate an immigration from the occurrence of two biomes. This is partly due to the central Europe: e.g. Discus ruderatus, Clausilia geomorphology of Europe, notably the northern plain pumila, Helicigona striata, Acicula polita, Macrogastra and the central Alpine mountains, which allows a ventricosa, Truncatellina callicratis, Vitrinobrachium dispersal from east to west for taxa living in periglacial breve, Ruthenica filograna (Kerney, 1977; Holyoak and environments (Fig. 2). Preece, 1986). Laminifera pauli which now lives in the In the interglacials, the topography of southern region of Bayonne (Pyrrnres atlantiques, France) has Europe limited potential immigration routes to only also been found in England (Kerney, 1959; Holyoak et three, the Atlantic littoral, the Sarne-Rh6ne valley al., 1983). But the lack of some taxa, which have a and the Danubian corridor (Fig. 2). The use of the south-western European distribution today, in the latter by the H. banatica fauna has been demonstrated interglacial fauna in western Europe was surprising to by Lozek (1964, 1969). Up to now, the emigration Lozek (1969). along the Rhodanian corridor has not been given Why did this mixing of the interglacial malacofauna, prominence because the interglacial faunas found in with the occurrence of 'foreign taxa', only take place in Burgundy and Alsace do not reveal the occurrence of central Europe, although favourable environmental southern immigrants (Puissrgur, 1976, 1978). On the conditions existed in western Europe (Fig. 2)? This other hand, revisions of the malacofauna of Normandy reveals on the one hand, the lack of complementary show interglacial taxa with a clear southern distribu- data about France and England at the time of Lozek's tion. (1969) review, and on the other hand, conceptual In western Europe, during the 'Holsteinian' or the differences, previously mentioned, about the British 'Hoxnian', tufa sedimentation developed in a palaeobiogeographical processes. moist forest context under interglacial conditions. Such Causal Ecological Biogeography 597 deposits occur at: Hitchin and Icklingham (England); Saint-Pierre-l~s-Elbeuf, Vernon and Arrest (France). There are some doubts about the age of the latter, because correlation is based on the malacofauna only (Fig. 2). Another site which might also be included is La Celle-sous-Moret, but the collection of Munier- Chalmas has been lost. Besides the common forest species typical of deciduous temperate forest (determined by the forest and semi-forest species sensu; Puiss6gur, 1976), these deposits yield several taxa with a distinctive recent distribution (Puiss6gur, 1983).

Botanical data The site of La Celle-sous-Moret has provided plant macro-fossils and pollen, among which occur Laurus canariensis (L. azorica), Canary laurel Ficus carica (fig), Cercis siliquastrum (Judas tree), Buxus semper- FIG. 4. Present distribution (shading) of Acicula polita in northern virens (box), Cerasus mahaleb (the St Lucy cherry), and central Europe (modifiedfrom Kerney et al., 1983). The scale indicates 1 mm high. and Viburnum tinus (Viburnum). All these taxa are now Mediterranean, with Laurus canariensis now being confirmed to the Canaries (Bourdier, 1969; Guyot, 1969). In the Vernon tufa, Bourdier (1969) mentions the occurrence of the same species, which, with the malacofauna allows him to correlate the two sites.

Malacological data The following are species peculiar to each deposit. Icklingham-Hitchin: (from Kerney, 1976; Holyoak et al., 1983; Holyoak and Preece, 1986): --Acicula diluviana (fossil); Retinella (Lyrodiscus) skertchlyi (fossil) (a); Ruthenica filograna (currently in central Europe; Fig. 3); Acicula polita (currently in central Europe; Fig. 4); -- Laminifera pauli (Pyrenean, around Bayonne at present; Fig. 5). Arrest (Puiss6gur, unpublished data; Rousseau and FIG. 5. Presentdistribution (shading)of Laminifera pauli in northern and central Europe (modifiedfrom Kerney et al., 1983). The scale Puiss6gur, 1990): indicates 1 mm high. -- Retinella (Lyrodiscus) skertchlyi (fossil) (a).

Vernon (Puiss6gur, unpublished data; Rousseau and Puiss6gur, 1989, 1990): --Aegopinella bourdieri (fossil) (b); Retinella (Lyrodiscus) skertchlyi (fossil) (a); --Hygromia cinctella (currently Mediterranean; Fig. 6); -- Leiostyla anglica (currently British Isles; Fig. 7); -- Ruthenica filograna (currently central Europe); --Zonites acieformis (currently central Europe; Fig. 8). Saint-Pierre-16s-Elbeuf (Puiss6gur, 1983; Rousseau and Puiss6gur, 1990): -- AegopineUa bourdieri (fossil) (b); -- Bradybaena chouquetiana (fossil); -- Retinella (Lyrodiscus) skertchlyi (fossil) (a); -- Hygromia cinctella (currently Mediterranean); --Laminifera pauli currently Pyrenean, near FIG. 3. Present distribution (shading) of Ruthenica filograna in northern and central Europe (modified from Kerney et al., 1983). Bayonne); The scale indicates 1 mm high. -- Zonites acieformis (currently central Europe). 598 D.-D. Rousseau

(a) Retinella (Lyrodiscus) skertchlyi is a zonitid described by Kerney (1976) from the tufa of Ickling- ham. This taxon is very interesting. It had a west- European distribution during the Plio-Pleistocene, but its related recent species are confirmed to the Canary Islands (Rousseau and Puissrgur, 1990). So its occurrence, in deposits with the Canary laurel is an important biogeographical record. (b) Aegopinella bourdieri is a new fossil species which has so far only been found in Normandy, at Vernon, and St Pierre-l~s-Elbeuf (Rousseau and Puis- srgur, 1989). The association of recent forest and semi-forest taxa, together with 'immigrant' species which have changed their distribution and their ecological peculiarities, allows us to characterize a peculiar faunal community. FIG. 6. Present distribution (shading) of Hygromia cinctella in Taking into account the botanical data which indicate northern and central Europe (modified from Kerney et al., 1983). The scale indicates 1 mm high. the same type of association, the fauna and flora of these tufas constitute an interesting biogeographical association. The environment in which these fossils are found is similar for each deposit, a spring tufa produced by seepage of calcareous water. Such analogies contri- bute to characterize a biome, a grouping together of a plant association of temperate deciduous forest with xerophilous taxa, and a particular malacological association with southern species, which reached its climax during tufa sedimentation. The currently known distri- butional area of the biome is from west Suffolk in England to Normandy in France.

DISCUSSION AND CONCLUSION

Do such faunal and floral associations have any links? According to the evidence of present faunas, the heterogeneous nature of the botanical and malacologi- cai associations is the limiting factor for any interpretation. No a priori argument allows us to assert or to FIG. 7. Present distribution (shading) of Leiostyla anglica in northern and central Europe (modified from Kerney et al., 1983). The scale invalidate the idea that these emigrant taxa have kept indicates 1 mm high. or changed their ecological affinities and peculiarities. Nevertheless, we have to consider such a biome with causal biogeography.

The Composition of the Biome Plant and animal species reveal a temperate open forest. But the immigrant taxa yield special informa- tion (Becker et al., 1982). Buxus sempervirens (Fig. 9) is mainly common in the south of France. It is tolerant of drought caused by summer heat, but is also hardy in winter. It is a common plant in the Holsteinian and Eemian in northwestern Europe. Cerasus malaheb is a species of the hills and lower slopes of southern mountains, requiring some heat but less than the typically Mediterranean species. It is tolerant of dry- ness, but requires good light. Cercis siliquastrum (Fig. 10) is spontaneous in the Mediterranean region; it lives on dry calcareous hillsides, never in forest, but in open vegetation, and is moderately hardy against the cold. Ficus carica (Fig. 11) lives in parts of the Mediterra- FIG. 8. Present distribution (shading) of Aegopis verticillus in northern and central Europe (modified from Kerney et al., 1983). nean, in warm and sunny, often, rocky localities. The scale indicates 1 mm high. Viburnum tinus, lives in the garrigue and in the scrub, Causal Ecological Biogeography 599

FIG. 9. Present distribution (shading) of Buxus sempervirens in FIG. 11. Present distribution (shading) of Cercis siliquastrum in western Europe (modified from Becker et al., 1982). The scale western Europe (modified from Becker et al., 1982). The scale indicates 5 cm high. indicates 5 cm high.

The Origin of that Biome As indicated above, such a forest type developed during the Tertiary in Europe and in NW Africa (Raven and Axelrod, 1974). From that time the laurel forest changed progressively into a mixed forest. This change has been identified in Europe and in Africa (Raven and Axelrod, 1974). Some localities such as the Canary Islands currently show relics of that Tertiary forest. The occurrence in northwestern Europe of such a biological association during the Middle Pleistocene does not illustrate evolution on the spot, but an emigration from refugia permitted by favourable environmental conditions.

Complementary Elements Other sites of Hoxnian age have provided data which strengthen this interpretation. In England, Godwin (1977) indicates that the Hoxnian interglacial flora is composed of plants already known in England as well as in Ireland, but also Lusitanian or Hiberno-American species. The association is an open mixed-oak forest. FIG. 10. Present distribution (shading) of Ficus carica in western Europe (modified from Becker et al., 1982). The scale indicates 5 cm In southwest France, Texier et al. (1983) indicate high. during their 'Mindel-Riss' stage, the occurrence of a flora of numerous thermophilous taxa (mixed-oak) and hygrophilous herbs. High temperature conditions are but never in the drier localities. Finally, Laurus indicated by the occurrence of Mediterranean species canariensis inhabits the Canary Islands laurel forest, (i.e. Rhododendron ponticum, Erica mediterranea) and which "is a remnant of the much richer evergreen forest by taxa with Lusitanian affinities (Dabaoecia canta- that covered parts of Europe and Africa in the brica). Tertiary" (Raven and Axelrod, 1974). The deposition of tufa requires the conjunction of All the mollusc species mentioned inhabit forest or different influences in which the local topography, the semi-forest habitats. So the vegetation is a temperate, environment, the possibility of the mobilization of non tropical, sparse mixed forest which has existed in calcium carbonate in the vicinity, and also the climate western Europe since the Late Miocene and in the are important. Because the southern taxa seem to have northwestern Mediterranean region since the Reuve- kept their ecological character, the development of rian (Suc and Zagwijn, 1983; Suc, 1985). such a deposit implies the occurrence of a warmer 600 D.-D. Rousseau climate than that recently observed in Normandy or to the detriment of the temporal factor. In other terms, England. the distinction of the biome and the biota concepts leads to an inversion of the balance of the ecological The Climate and of the temporal factors. As the integration of both In all cases the climate indicated is highly oceanic botanical and zoological data allows us to better delimit with high summer temperatures. Such a climate is the real conditions of life and thus lead to a more unlike the present one in this region. It seems difficult precise research orientation, causal ecological biogeo- to consider such a biome as the result of a special graphy is certainly progressive. By its hypothetico- microclimate, so changes in the oceanic and atmos- deductive process, this approach is a successful re- pheric circulations may have occurred. The sites are search programme which involves channeling located in Normandy and also in eastern England. So, approaches to a problem. From such data, only few if we consider a modification of the oceanic circulation, appropriate hypotheses can be developed which will be this might be appropriate for Normandy and also for confirmed or denied by new studies. The analysis of the southwest France. But for west Suffolk, it would be biome of the Holsteinian tufas of northwestern Europe necessary to propose warm waters in the North Sea. has been proposed to demonstrate this point of view. What is the age of all these tufas? The ages of So the multidisciplinary implications of this analysis in Hoxnian and of the Hoisteinian are the subjects of the Life Sciences as in Earth Sciences indicates that this debate. Amino acid dating based on land snail shells by scientific research programme, up to now undeveloped, Occhietti (unpublished data) indicate a younger age for is particularly appropriate for Quaternary palaeontolo- the tufa of St Pierre-l~s-Elbeuf than that proposed by gy, and seems more suitable for further development. Lautridou. The hypothesis of Lautridou, the immigrant episode in western Europe, is younger than that of central Europe. Following his interpretation, and using ACKNOWLEDGEMENTS the nomenclature of Kukla, the St Pierre tufa corresponds to the interglacial of glacial Stage E. In central Special thanks to Drs J.L. Dommergues (Dijon), D.H. Keen (Coventry), and W.H. Zagwijn (Haarlem), for discussion and for Europe H. banatica faunas occur in the interglacials of reading the manuscript. Thanks to Dr J.P. Lautridou (Caen) for field the glacial Stages B and D (Kukla, 1977). But are these discussion and to Dr J.J. Puissrgur (Dijon) who allowed me to study immigrant episodes synchronous? Do they correspond part of his collection. This work was supported by CNRS. to a response to the same physical constraints but of different ages? Where do the immigrant taxa come REFERENCES from? More data are necessary to establish the distribution of these taxa during both interglacials and glacials. Aho, J. (1984). Relative importance of hydrochemical and equilibrial Two main conclusions are possible. First, the variables on the diversity of freshwater gastropods in Finland. In: development of that biome raises again the problem of Solem, A. and van Bruggen, A.C. (eds), World-Wide Snails. Biogeographical Studies on Non-Marine Mollusca, pp. 198-206. the correct stratigraphical position of the tufa of St Brill and Backhuys, Leiden. Pierre-16s-Elbeuf. Is it part of a pedocomplex overlying Andrr, J. (1984). Biogeographical studies on the molluscs of the the Elbeuf II palaeosol, or are they distinct? At bioclimatological region of the Mediterranean parts of the Iberian Peninsula and France. In: Solem, A. and van Bruggen, A.C. (eds), present, it is not possible to confirm either of these World-Wide Snails. Biogeographical Studies on Non-Marine possibilities without the help of complementary studies Mollusca, pp. 207-223. Brill and Backhuys, Leiden. such as micromorphology. Nevertheless, the examina- Becker, M., Picard, J.F. and Timbal, J. (1982). Larousse des Arbres, des Arbustes et des Arbrisseaux de I'Europe Occidentale. Librairie tion of the section does not allow us a priori to Larousse, Paris, 330pp. distinguish any erosional break between the tufa and Blondel, J. (1986). Biog~ographie Evolutive. Masson, Paris, 221pp. the palaeosol. If the hypothesis of the pedocomplex Bourdier, F. (1969). Etude comparEe des drprts quaternaires des bassins de la Seine et de la Somme. Bulletin d'lnformation des was confirmed, two interglacial biomes would have G~ologues du Bassin de Paris, 21, 169-220. occurred during the same period in western and central Clements, F.E. (1949). Dynamics of the Vegetation. Hafner. New Europe, which show malacological communities with York. Craw, R.C. and Weston, P. (1984). Panbiogeography: a progressive some southern emigrant species. For this episode, research program? Systematic Zoology, 33 (1), 1-13. such a case would involve the occurrence of different Croizat, L. (1958). Panbiogeography or an introductory synthesis of atmospheric circulations providing a warm oceanic zoogeography, phytogeography and geology. 3 Volumes, (pub- lished by the author), Caracas, 1-1018 p., 2-1731p. temperate climate in western Europe and a more Dommergues, J.L. and Marchand, D. (1988). Pal~.obiogrographie Mediterranean continental temperate climate in central historique et 6cologique. Application aux ammonites du Juras- Europe reaching higher latitudes than at present. sique. In: Wiedmann, J. and Kullmann, J. (eds), 2nd International Cephalopod Symposium. Cephalopods: Present and Past, pp. 351- Second, concerning the biogeographical relationship, 364. Tiibingen Schweizerbartsche Verlag, Stuttgart. the biome concept characterizes a biogeographical Gittenberger, E. (1984). Vicariantists and dispersalists among the entity grouping the botanical and the zoological data Chondrininae (Gastropoda, Pulmonata). In: Solem, A. and van Bruggen, A.C. (eds), World-Wide Snails. Biogeographical Studies together in a fluctuating spatio-temporal scope. on Non-Marine Mollusca, pp. 56--69. Brill and Backhuys, Leiden. According to its nature, it does not correspond to the Godwin, H. (1977). Quaternary history of the British flora. In: purely stratigraphic biozone. Contrary to the biota Shotton, F.W. (ed.), British Quaternary Studies. Recent Advances, pp. 105-118. Clarendon Press, Oxford. concept, another fundamental biogeographical unit, Guyot, L. (1969). Histoire de la FIore Terrestre. Presses universitaires the biome, optimizes the weight of the ecological factor de France, Paris, 127pp. Causal Ecological Biogeography 601

Holyoak, D.T. and Preece, R.C. (1986). A reassessment of some PuissEgur, J.J. (1978). Les mollusques des series loessiques British Pleistocene land snails. Journal of Conchology, 32, Achenheim. Recherches Gdographiques ~ Strasbourg, 7, 71-96. 199-200. PuissEgur, J.J. (1983). Faunes malacologiques plEistocEnes de Holyoak, D.T., Ivanovich, M. and Preece, R.C. (1983). Additional Normandie. In: Lautridou, J.P. (ed.), version franqaise de 'The fossil and isotopic evidence for the age of the interglacial tufas at Quaternary of Normandy, Guide-Book of the Q.R.A. Normandy Hitchin and Icklingham. Journal of Conchology, 31, 260-261. meeting, May 1982 and IGCP 24 Normandy meeting, September Janvier, Ph. (1982). La biogEographie phylogEnEtique et la vicari- 1982,' pp. 151-160. DEpartement de GEologie UER Sciences, ance: historique et perspectives d'avenir. Bulletin de la Soci~t~ Rouen. G~ologique de France, Paris, (7) t.29, 5-6, 917-925. Raven, P.H. and Axeirod, D.I. (1974). Angiosperm biogeography Janvier, Ph. (1986). Comment 6valuer les liens entre structures and past continental movements. Annals of the Missouri Botanical biogEographiques et Ev~nements gEologiques? Une analyse Garden, 61, 539-673. critique des mEthodes en biogEographie et palEobiogEographie. Rosen, D.E. (1978). Vicariant patterns and historical explanation in Bulletin des Centres de Recherche et d'Exploration-Production biogeography. Systematic Zoology, 27, 159-188. d'Elf-Aquitaine, Pau, 10 (2), 323-332. Rousseau, D.D. (1986). IntErEt palEobiogEographique de Pupilla Kerney, M.P. (1959). An interglacial tufa near Hitchin Hertford- loessica (Lozek) et de Vallonia tenuilabris (A. Braun) pour le shire. Proceedings Geologists Association, 70, 322-337. PlEistocEne Ouest-europEen. Comptes Rendus de l'Acad~mie des Kerney, M.P. (1976). Mollusca from an interglacial tufa in East Sciences, 3tl3, SEr. II, 3, 257-262. Anglia, with the description of a new species of Lyrodiscus Pilsbry Rousseau, D.D. (1987). Paleoclimatology of the Achenheim series (Gastropoda: Zonitidae). Journal of Conchology, London, 29, (Middle and Upper Pleistocene, Alsace, France). A malacologi- 47-50. cal analysis. Palaeogeography, Palaeoclimatology, Palaeoecology, Kerney, M.P. (1977). British Quaternary non-marine Mollusca: a Amsterdam, 59, 293-314. brief review. In: Shotton, F.W. (ed.), British Quaternary Studies. Rousseau, D.D. and PuissEgur, J.J. (1989). Aegopinella bourdieri Recent Advances, pp. 31-42. Clarendon Press, Oxford. nov. sp. (Gastropoda: Zonitidae), une espEce nouvelle du PlEis- Kerney, M.P., Cameron, R.A.D. and Jungbluth, J.H. (1983). Die tocEne europEen. Geobios, 22, 123-128. Landschnecken Nord- und Mittleeuropas. Paul Parey, Hamburg, Rousseau, D.D. and PuissEgur, J.J. (1990). PhylogenEse et biogEo- 384pp. graphie de Retinella (Lyrodiscus) Pisbry (Oastropoda; Zonitidae). Kukla, G.J. (1977). Pleistocene land-sea correlations. 1. Europe. Geobios, 23, 57-70. Earth-Sciences Review, Amsterdam, 13, 307-374. Shikov, V. (1984). Effects of land use changes on the land mollusc Lautridou, J.P., SommE, J., Heim, J., PuissEgur, J.J. and Rousseau, fauna in central portion of the Russian plain. In: Solem, A. and D.D. (1985). La stratigraphie des loess et formations fluviatiles van Bruggen, A.C. (eds), World-Wide Snails. Biogeographical d'Achenheim (Alsace): nouvelles donnEes bioclimatiques et Studies on Non-Marine Mollusca, pp. 237-248. Brill and Backhuys, corr61ations avec les s6quences plEistocEnes de la France du Nord- Leiden. Ouest. Bulletin de l'Association Fran~aise pour I'Etude du Quater- Schlickum, W.R. (1975). Die oberplioz/ine MoUuskenfauna von naire, Paris, 22--23, 125-132. Cessey-sur-Tille (D6partement CEte-d'Or). Archiv fiir Mollus- Lautridou, J.P., SommE, J., Heim, J., Maucorps, J., PuissEgur, J.J., kenkunde, Frankfurt, 106 (1/3), 47-80. Rousseau, D.D., ThEvenin, A. and van Vliet-LanoE, B. (1986). Solem, A. (1984). A world model of land snail diversity and CorrElations entre sediments quaternaires continentaux et marins abundance. In: Solem, A. and van Brnggen, A.C. (eds), World- (littoraux et profonds) dans le domaine France septentrionale- Wide Snails. Biogeographical Studies on Non-Marine Mollusca, Manche. Revue de G~ologie Dynamique et G~ographie Physique, pp. 6-22. Brill and Backhuys, Leiden. Paris, 27 (2), 105-112. Solem, A. and van Bruggen, A.C. (eds) (1984). World-Wide Snails. LEmEe, G. (1967). Precis de Biog~ographie. Masson, Paris, 358pp. Biogeographical Studies on Non-Marine Mollusca. Brill and Lozek, V. (1964). Quart~irmollusken der Tschechoslowakei. Backhuys, Leiden, 289pp. Rozprayy Ustredn~o tlstavu geolgickEho, Prague, 31, 374pp. Suc, J.P. (1985). Le climat mEditerranEen: une particularit6 de Lozek, V. (1969). Le loess et les formations assimilEes: correlation toujours? La Recherche, 162, 87-89. entre I'Europe centrale et la France par la faune de mollusques. In: Suc, J.P. and Zagwijn, W.H. (1983). Plio-Pleistocene correlations Etudes sur le Quaternaire dans le monde VIII Congas INQUA between the northwestern Mediterranean region and northwestern Paris, pp. 597-606. Europe according to recent biostratigraphic and palaeoclimatic Meier-Brook, C. (1984). A preliminary biogeography of freshwater data. Boreas, 12, 153-166. pulmonate gastropods. In: Solem, A. and van Bruggen, A.C. Texier, J.P., Raynal, J.P., Laville, H., Paquereau, M.M., Prat, F., (eds), World-Wide Snails. Biogeographical Studies on Non-Marine Debenath, A. and Delpech, F. (1983). Histoire climatique de Mollusca, pp. 23-37. Brill and Backhuys, Leiden. l'Aquitaine du PlEistocEne ancien ou dernier interglaciaire. Acres PuissEgur, J.J. (1976). Mollusques continentaux quaternaires de CoUoque AGSO Bordeaux Mai 1983. Bulletin de l'lnstitut de Bourgogne. Significations stratigraphiques et climatiques. Rap- G~ologie du Bassin d'Aquitaine, 34, 207-217. ports avec d'autres faunes borEales de France. M~moires G~o- logiques de l'Universit~ de Dijon, 3, 241pp.